Life Satisfaction, Self-Concept, and Family Relations in Chinese Adolescents and Children

Download as pdf or txt
Download as pdf or txt
You are on page 1of 8

International Journal of Behavioral Development # 2003 The International Society for the

2003, 27 (2), 182–189 Study of Behavioural Development


http://www.tandf.co.uk/journals/pp/01650254.html DOI: 10.1080/01650250244000182

Life satisfaction, self-concept, and family relations in


Chinese adolescents and children

Lei Chang and Sunita M. Stewart Ernest Au


Catherine McBride-Chang University of North Texas, Chinese University of Hong Kong,
Chinese University of Hong Kong, Dallas, USA Hong Kong
Hong Kong

Subjective well-being across the life span may be affected by both age-specific and age-general factors
within a cultural context. Thus, this study explored both developmentally invariant and variable
predictors of life satisfaction among 115 second-graders and 74 eighth-graders from Hong Kong. In a
regression model, general self-concept and ratings of parental warmth and autonomy/detachment
predicted life satisfaction equally across the two age groups. However, social self-concept was a
strong predictor of life satisfaction among adolescents only, whereas actual academic test scores
predicted life satisfaction only among the children. Mean group differences emerged as well, with
adolescents scoring significantly lower in life satisfaction and self-concept and higher in emotional
detachment than children. Results are explained in relation to both development and culture.

Introduction tions derived from these two domains of life. Among the
hypotheses, we expected parental warmth and general self-
In many ways, life satisfaction is the ultimate goal of human concept to invariably predict, and social self-concept, academic
development. Competent cognitive and academic develop- performance, and emotional detachment from parents to
ment can be seen as a means to achieving life satisfaction, differentially predict life satisfaction of these two age groups.
whereas social emotional adjustment is synonymous with A second and related objective of the present study was to
subjective well-being. Much research has been conducted to examine developmental change or stability in life satisfaction
identify factors leading to or hindering the attainment of life and self-concepts in specific domains within the cultural
satisfaction in both Western (e.g., Dew & Huebner, 1994; context of these two groups of Chinese children.
Heaven, 1989) and Chinese adolescents (J.-P. Leung & Zhang,
2000; J.-P. Leung & Leung, 1992). However, of greater
developmental importance is determining the degree to which
Self-concept and life satisfaction
life satisfaction is similarly or differentially perceived by Self-concept and life satisfaction are related to the same self-
individuals across stages of development. The Eriksonian representation system. Life satisfaction reflects the affect
concept of development in part emanates from observed shifts dimension of the self-system (Petersen, Compas, Brooks-
in individuals’ interests in and concerns over different tasks Gunn, Stemmler, Ey, & Grant, 1993), whereas self-concept
and activities faced in different phases of life (Erikson, 1968). represents cognitive appraisals of the self ’s competencies and
Individuals, thus, possess internal representations of their life weaknesses (Terry & Huebner, 1995). Both systems are ideal
priorities that allow them to derive life satisfaction throughout for the study of developmental transitions, e.g., from childhood
development. Some internal representations are more or less to adolescence, as the latter especially affects individuals’ self-
related to life satisfaction, depending on the order in which the perceptions (Harter, 1990). Within the framework of multi-
underlying life events are prioritised at a given age. Other dimensional self-conceptualisation (Harter, 1986), domain-
representations may be invariant across the developmental specific self-concepts may follow similar or variable develop-
span as constant sources of life satisfaction. mental trajectories depending on the developmental relevance
The purpose of the present study was to identify both and valence of a given domain. Depending on the domain
developmentally invariant and different self-representations in relevance at a given developmental time, different domain-
predicting life satisfaction across two phases of child develop- specific self-appraisals (or self-concepts) may similarly or
ment. In two cross-sectional samples of 115 primary school differentially affect individuals’ subjective affect such as life
(Grade 2) children and 74 early adolescents (secondary school satisfaction. Similarly, the association of a domain-specific self-
Grade 2 or North American Grade 8) in Hong Kong, we concept with life satisfaction may either fade or strengthen
looked for invariant and different predictors of life satisfaction from childhood to adolescence.
according to life priorities of these children within their cultural The domain of friendship and social interaction is believed
context. For both age groups, family and school constitute to be more relevant during adolescence than childhood. Early
major life activities. Thus, we examined internal representa- adolescence has been identified as a developmental period

Correspondence should be addressed to Lei Chang, Department of (No. EMB/QEF/P1999/2735) from the Hong Kong SAR government
Educational Psychology, Chinese University of Hong Kong, New to the first and second author.
Territories, Hong Kong; e-mail: leichang@cuhk.edu.hk. Thanks are given to the students, parents, and teachers of the two
This study was supported by a Quality of Education Fund grant participating schools of this project.

Downloaded from jbd.sagepub.com at MOUNT ALLISON UNIV on June 21, 2015


INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2003, 27 (2), 182–189 183

during which the salience of peer relations increases as the both the West and East, sport is an increasingly important
salience of the family decreases (Laursen, 1996; Steinberg & aspect of adolescents’ lives, with athletic abilities being
Silverberg, 1986). Coinciding with the onset of puberty, this is significant determinants of peer status (e.g., Dong, Weisfeld,
also a period when adolescents begin to value friendship as an Boardway, & Shen, 1996). Due to scarcity of space, many
affective necessity (Richards, Crowe, Larson, & Swarr, 1998). schools in Hong Kong have small sports fields, if they have any.
The nature and structure of social interaction also change in Except for horse racing, commercial sport clubs and competi-
late childhood and adolescence. In the early grades of primary tions are almost nonexistent. There are few organised
schools, friends are chosen based on convenience (Selman, tournaments or regular sporting events that are participated
Levitt, & Schultz, 1997). In contrast, middle school adoles- in by local athletes. Nor is there a system for selecting and
cents begin to make conscious efforts in apprenticing their training potential athletes, either self-sustained in Hong Kong
social skills, exerting social influence, and establishing peer or linked to Mainland China. Thus, in comparison to many
status. In response, social self-concept in the middle school other societies and cultures, youth in Hong Kong may lack an
years affects the subjective affect of adolescents more than that athletic channel through which to develop their potential. This
of primary school children. Several empirical studies lend ecological characteristic provides an opportunity to examine
support to this hypothesis. For example, Larson and Richards domain relevance in regulating the relation between self-
(1991) reported that middle school children spend more time concept and subjective affect.
with friends and less with family than primary school children, Finally, general self-concept is hypothesised to predict life
and time with friends was a more important determinant of satisfaction equally for adolescents and primary school
happiness in adolescence. In younger children, family relations children. This is consistent with earlier theorising that self-
were more strongly associated with life satisfaction than concept differentially predicts life satisfaction because the
friendship relations (Huebner, 1991). Given this developmen- valence of the underlying domain relevance changes with
tal characteristic, social self-concept was hypothesised to development. Empirical evidence also supports invariance of
predict life satisfaction of adolescents more strongly than that general self-concept in predicting life satisfaction between
of primary school children. adolescents (Heaven, 1989; J.-P. Leung & Zhang, 2000) and
Academic activities, on the other hand, represent a domain younger children (Terry & Huebner, 1995). Also of develop-
that is highly relevant for both primary and secondary school mental importance, it is a study of normative change of
children in Hong Kong. A strong emphasis on academic different self-concepts across developmental transitions. Past
achievement in Chinese children has long cultural roots. research suggests that, across many domains, early adolescents
However, perhaps even stronger than these distal influences first entering middle schools experience a significant drop in
are the limited opportunities in tertiary education of self-concept, which is rebuilt during late adolescence (see
contemporary China and other East Asian countries that Harter, 1998, for a review). On the other hand, children in
place children into an indefatigable and ceaseless competition early primary school have an inflated self-concept (Eccles,
from kindergarten into high school. The eight universities Midgley, & Adler, 1984; Wigfield, 1994). They tend to be
operating in Hong Kong, for example, absorb only 18% of unrealistically optimistic (Eccles et al., 1984). This age
the middle school student population, which numbers 75,000 difference in self-concept in part also reflects transitional stress
each year. In this academically competitive environment, self- experienced by early adolescents. Early adolescence coincides
concept of cognitive and academic abilities in both children with simultaneous transitions of three kinds. There is the
and adolescents is expected to be highly relevant to their life educational transition from primary school to middle school,
satisfaction. cognitive emergence of formal operation reasoning, and, in
However, the actual academic performance as indicated by most Chinese children, onset of puberty. The cumulative
test scores was hypothesised to be more predictive of life effects of these transitions result in heightened stress in this age
satisfaction in primary school children than adolescents. This period (e.g., see Eccles & Midgley, 1989, for a review) that
hypothesis is based on the observation that adolescents’ self- often leads to depressed self-perceptions (Eccles, Wigfield,
system is more dynamic and multidimensional than younger Midgley, Reuman, MacIver, & Feldlaufer, 1993; Harter, 1990;
children’s simply because adolescents have wider life experi- Wigfield, Eccles, MacIver, Reuman & Midgley, 1991). The
ences than younger children (Eccles, Wigfield, Flanagan, affect dimension of the self-system also experiences depression
Miller, Reuman, & Yee, 1989). As children grow older and during this transition (e.g., Morganti, Nerke, Hulicka, &
spend more time in schools, competence-related feedback Cataldo, 1988; Petersen et al., 1993). A 50% decline has been
becomes more frequent and more variable. Reliance on test reported in ninth-graders as compared to fifth-graders in self-
scores as the sole barometer of abilities is complemented and reports of positive feelings such as ‘‘great’’ and ‘‘very happy’’,
refined by other more implicit means of comparisons. The supporting the view that adolescence is a time of ‘‘deflation of
education system in Hong Kong also deems test scores to be childhood happiness’’ (p. 85, cited in Arnett, 1999). Given this
less relevant indicators of abilities in middle schools than in literature, the four domains of self-concept and life satisfaction
primary schools. Middle schools in Hong Kong are streamed were expected to be lower in adolescents than in primary
by primary school test scores into bands. (More on Hong Kong school children.
school systems in the Method section.) Students within a
secondary school have similar test scores. Thus, test scores that
are more differentiating in primary than secondary schools
Family relations and life satisfaction
were hypothesised to predict primary school children’s life In Chinese societies, children are inculcated with strong family
satisfaction more strongly than that of adolescents. values and the belief that family and kinship relations are
Another culturally interesting phenomenon related to fundamental to happiness in life (Ho, 1986). This socialisation
children’s domains of activities in this study is the surprisingly tenet is present throughout childhood and adulthood. In
low emphasis on sports in Hong Kong. In most countries in Chinese societies, filial piety and its reciprocal parental

Downloaded from jbd.sagepub.com at MOUNT ALLISON UNIV on June 21, 2015


184 CHANG ET AL. / LIFE SATISFACTION IN YOUNG CHINESE

commitment to children’s well-being idealise not only young school children, 113 lived in two-parent households, while 73
children’s parent–child relations but adolescents’ and adults’ as of the adolescents lived with both parents.
well. In Hong Kong, which had been a British colony for 150
years until 1997, this Chinese tradition is also juxtaposed with
Christian values of love within a nuclear family. In the Western Procedure
literature, parental warmth has been shown to have an Following parental consent, questionnaire data were obtained
overarching positive effect leading to a wide range of positive from both students and their parents. Primary school students
experiences in children of different ages (e.g., see Cournoyer, were allotted a single class period in which to complete all
2000, for a review). According to Kagitcibasi (1996), parental questionnaires. Four primary school classes participated. In
warmth is a universal contributor to well-being among children each classroom, questions were read aloud by an under-
and adolescents. In a number of Chinese studies, parental graduate psychology major while children rated each item.
warmth has similarly been associated with positive subjective Four other psychology majors monitored the children’s
well-being in both children (Stewart, Rao, Bond, McBride- progress throughout the questionnaire completions. They were
Chang, Fielding, & Kennard, 1998) and adolescents (J.-P. available to answer questions about individual items, reread
Leung & Zhang, 2000; Yau & Smetana, 1996). Given the questions, and ensure that students fill out the questionnaires
empirical literature as well as cultural theorising, a close at an appropriate pace.
parent–child relationship as indicated by parental warmth was Secondary school adolescents completed the questionnaires
expected to be an invariant contributor to children’s and after school. All volunteers came to one of two after-school
adolescents’ life satisfaction. sessions and completed the questionnaires individually. Five
In addition to parental warmth, we also examined Steinberg undergraduate psychology majors were on hand to answer
and Silverberg’s (1986) measure of autonomy/detachment (A/ questions about the questionnaires when needed. Both the
D). McBride-Chang and Chang (1998) administered A/D to primary and secondary students received McDonald’s coupons
Hong Kong youth. They noted an overall increase in A/D in appreciation for their participation.
among older adolescents relative to younger ones, as found
previously in American adolescents (Steinberg & Silverberg,
1986). In several Western studies (e.g., Fuhrman & Holmbeck, Demographics
1995; Ryan & Lynch, 1989; Smetana, 1995), the A/D scale
tended to be negatively associated with positive parenting One parent completed all demographic information for the
measures (e.g., family cohesion, maternal warmth, authorita- present study. The questionnaire included information on both
tive parenting). Thus, this scale has been conceptualised as a parents’ education levels and ages, as well as parents’ marital
measure of emotional detachment from parents (McBride- status, the numbers of children and adults in the family, and
Chang & Chang, 1998; Ryan & Lynch, 1989). The A/D scale fathers’ income levels. Education level was measured on a 6-
has not been previously administered to children. Its inclusion point scale ranging from 1 ¼ primary school to 6 ¼ graduate
in the present study allows the examination of the development school. Father’s income was assessed on a 5-point scale
of emotional detachment and its psychological impact over ranging from 1 ¼ HK$6000 to 5 ¼ more than HK$40,000 per
age. We predicted that adolescents would score higher on this month (Hong Kong dollars are pegged to US dollars at a fixed
measure than would children based on previous studies rate of 7.8). These demographic statistics are given in Table 1.
demonstrating increased A/D with age in adolescence Because the adolescent group was 6 years older than the child
(McBride-Chang & Chang, 1998; Steinberg & Silverberg, group, ages of students, mothers, and fathers of these groups
1986). Because it has not been used in children, a specific age differed significantly. Furthermore, there were predictably
effect on its relationship with life satisfaction is unclear. more children in the families of the adolescents than in those
Because of its negative associations both with positive of the children, whose families were younger. However, there
parenting measures (e.g., parental warmth, Fuhrman & were no significant differences in the samples on parents’
Holmbeck, 1995; Ryan & Lynch, 1989; Smetana, 1995) and education levels; most parents had completed secondary
with psychological adjustment (Beyers & Goossens, 1999), we school, and many had had some college training as well. In
expected it to affect life satisfaction negatively. addition, across groups, the numbers of adults in the families
and fathers’ income levels were not significantly different,
suggesting that the groups of children and adolescents were
adequately matched.
Method
Participants Measures
Participants were 115 second-graders (65 girls, 50 boys) from a All the measures have been translated into Chinese by other
single primary school in Hong Kong, 74 second-graders researchers or the present research team and have been
(equivalent to North American eighth-graders; 39 girls, 35 checked by an independent bilingual person for truthfulness
boys) from a secondary school, and both of their parents. of meaning and by teachers of the participating students for
Secondary schools in Hong Kong are segregated according to a appropriate language use. As presented below, all the instru-
banding system. Based on various primary school test results, ments have been used previously with Chinese participants of
students are placed in secondary schools of one of five bands similar age and background.
(from the 2002 school admission, this will be changed to three Life satisfaction was measured by Huebner’s (1994) 40-item
bands), with Band 1 being the highest and Band 5 being the questionnaire, the Multidimensional Students’ Life Satisfac-
lowest in terms of academic performance. The present tion Scale. Examples of items are ‘‘I look forward to going to
secondary school was a Band 2 school. Among the primary school’’ and ‘‘I wish I lived somewhere else’’. The instrument

Downloaded from jbd.sagepub.com at MOUNT ALLISON UNIV on June 21, 2015


INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2003, 27 (2), 182–189 185

Table 1
Comparisons of adolescents and children on demographic variables

Adolescents Children

Variable Mean (SD) Mean (SD) T-value

Age 13.6 (0.6) 7.8 (0.5) 71.4***


Mother’s age 40.5 (7.2) 37.7 (5.1) 3.1**
Father’s age 45.5 (5.6) 41.4 (6.6) 4.3***
Mother’s education 2.9 (1.0) 3.2 (1.3) 1.6
Father’s education 3.1 (1.2) 3.5 (1.6) 1.8
Father’s income 2.7 (1.1) 2.9 (1.1) 1.1
Number of children 2.3 (0.9) 2.0 (0.7) 2.4*
Number of adults 2.7 (1.2) 2.7 (0.9) 0.1

*p 5 .01; **p 5 .005; ***p 5 .001.

has been previously used with Hong Kong adolescents (Y.W. sample for academic, social, sports, and general, respectively,
Leung, 2001). The questionnaire can be used as five subscales were .80, .81, .76, and .75.
measuring satisfaction within different domains or as a single Test scores were obtained from official school records. Scores
scale of overall life satisfaction. A single scale was used in the in the three subject areas considered to be of utmost
present study. The obtained internal consistency reliability for importance by schools—Chinese, mathematics, and Eng-
this measure was .86 in this study. lish—were obtained. We combined the scores into a compo-
Autonomy/detachment was measured using a 14-item version site. Because the two schools used different scales in reporting
of Steinberg and Silverberg’s (1986) original Emotional these marks, within-school standardisation was taken to create
Autonomy Scale. This original scale consisted of four aspects a z-score composite to measure academic performance.
of autonomy–individuation, deidealisation of parents, nonde-
pendence on parents, and perceptions of parents as people
separate from their roles as parents. The last subscale, Results
perceptions of parents as people, has been demonstrated to
develop relatively slowly (Steinberg & Silverberg, 1986). Means, standard deviations, and correlation coefficients of all
Furthermore, in a study of Hong Kong youth (McBride- the variables used in the study are presented in Table 2. The
Chang & Chang, 1998), this dimension was stronger in same statistics are also reported for adolescents and primary
younger adolescents relative to older adolescents, in stark school children separately in Table 3.
contrast to the other three dimensions, which increased with
grade level, as predicted. For these reasons, the six-item
perceptions of parents as people subscale was not included in
Normative developmental differences
the autonomy/detachment scale. Items were presented on a 4- Multivariate analysis of variance was conducted to examine
point Likert scale ranging from ‘‘strongly agree’’ to ‘‘strongly the hypotheses that children scored higher on life satisfaction
disagree’’. Item analyses revealed that item 14 of the original and self-concept measures and lower on emotional detach-
scale, ‘‘My parents would be surprised to know what I’m like ment than adolescents. Also examined in the MANOVA was
when I’m not with them’’ loaded poorly with the other items gender as an additional factor. The multivariate F-test was
and was, therefore, not included in subsequent analyses using significant for age, F(6, 180) ¼ 22.85, p 5 .001, and for
the total score for this scale. This revised 13-item scale had an gender, F(6, 180) ¼ 6.02, p 5 .05, but not for a gender by age
internal consistency reliability of .73 in the present study. interaction. Univariate F tests suggest significant age differ-
Parental warmth was measured using a 5-item questionnaire ences in life satisfaction, F ¼ 9.51, p 5 .001, academic self-
for parents. Items included ‘‘I show my child through actions concept, F ¼ 17.76, p 5 .001, sports self-concept, F ¼ 10.59,
that I love him/her’’, ‘‘I tell my child (in words) that I love him/ p 5.001, and emotional detachment, F ¼ 18.07, p 5 .001.
her’’, and ‘‘I praise my child when I am pleased with his/her However, no significant difference was found in general self-
behaviour’’. Similar items have been used with Chinese parents concept and social self-concept. The significant mean differ-
elsewhere (Berndt, Cheung, Lau, Hau, & Lew, 1993). Internal ences are in the directions as hypothesised. That is, children
consistency reliabilities were .64 for mothers and .69 for fathers had higher means in life satisfaction and self-concept measures
in the present sample. Combining the two parents together but a lower mean in emotional detachment than adolescents.
resulted in much improved reliability of .82. The combined The means are reported in Table 3. Univariate tests revealed
scale was used as a measure of parental warmth. two gender effects. Boys scored statistically higher on sports
Self-concept was obtained by the Perceived Competence self-concept (male mean ¼ 18.97, female mean ¼ 17.73,
Scale for Children (Harter, 1982). The scale has four subscales t ¼ 2.12, p 5 .05) and slightly higher on emotional detachment
tapping domain-specific self-concept in cognitive and aca- (male mean ¼ 33.40, female mean ¼ 32.24, t ¼ 1.91, p 5 .06).
demic, social, physical and sports, and general self-worth. Each Additional analyses showed that girls had much higher
of these four subscales consists of seven items measured on a academic test scores than boys. These results are within
specially designed 4-point scale. (See Harter, 1982, for detailed expectations. For example, the Hong Kong Equal Opportu-
scaling and item information.) The scale has previously been nities Commission (1999) reported that girls consistently
used with Chinese children (Stigler, Smith, & Mao, 1985). outperformed boys in both primary and secondary school in
Internal consistency reliabilities estimated in the present Hong Kong. A similar gender difference has been widely

Downloaded from jbd.sagepub.com at MOUNT ALLISON UNIV on June 21, 2015


186 CHANG ET AL. / LIFE SATISFACTION IN YOUNG CHINESE

Table 2
Correlation coefficients, means, and standard deviations of the variables used in the study

1 2 3 4 5 6 7 8

1 Life satisfaction 1.00


2 Test scores .30 1.00
3 General self-concept .46 .12 1.00
4 Social self .34 .07 .56 1.00
5 Academic self-concept .38 .19 .61 .36 1.00
6 Sports self-concept .34 .11 .53 .54 .45 1.00
7 Parental warmth .32 .09 .23 .14 .28 .22 1.00
8 Emotional detachment .38 .02 .10 .06 .24 .17 .27 1.00
Mean 119.37 0.00 22.53 17.48 19.56 18.29 24.65 32.76
SD 14.96 1.00 3.66 3.22 3.80 4.05 2.74 4.18

Table 3
Correlation coefficients, means, and standard deviations of the variables in children and adolescents separately

1 2 3 4 5 6 7 8

1 Life satisfaction 1.00 .38 .39 .28 .28 .32 .29 .22
2 Test scores .18 1.00 .09 .10 .12 .11 .17 .07
3 General self-concept .57 .16 1.00 .64 .59 .65 .21 .04
4 Social self .47 .02 .44 1.00 .55 .65 .14 .08
5 Academic self-concept .43 .31 .64 .16 1.00 .64 .11 .01
6 Sports self-concept .28 .42 .32 .44 .13 1.00 .22 .03
7 Parental warmth .22 .01 .17 .18 .28 .07 1.00 .02
8 Emotional detachment .50 .05 .15 .13 .18 .04 .14 1.00
Adolescent Mean 115.08 0.00 21.95 17.58 18.16 17.09 23.40 35.93
SD 13.41 1.00 3.43 3.33 3.87 4.23 2.87 3.14
Child Mean 122.12 0.00 22.90 17.41 20.46 19.06 25.45 30.72
SD 15.31 1.00 3.77 3.16 3.49 3.75 2.34 3.44

Off diagonal are coefficients of adolescents, N ¼ 74, and above the diagonal are coefficients of children, N ¼ 115.

reported in the West as well (see Halpern, 1992 for a review). scores. Tested in the combined sample of children and
Because gender was not a focus of the present study, these adolescents, the results of this model are presented in
three gender differences will not be discussed further. Table 4. All the predictors were statistically significant except
sports self-concept and academic self-concept. Academic self-
concept was not significant when general self-concept was in
Age-related predictions of life satisfaction the equation. When general self-concept was not included, as
To test the hypotheses regarding age-related interaction effects reported in Table 4, academic self-concept was significant,
in the prediction of life satisfaction, multiple regression while statistically controlling other predictors. This finding
involving interaction variables was used (see Aiken & West, suggests that academic self-concept was an important part of
1991, for details). An interaction term was created for each of these children’s general self-concept and contributed to similar
the predictors of life satisfaction by multiplying school type, variations in life satisfaction.
which is a dummy variable representing primary versus The second model allowed social self-concept and test
secondary school, with the specific predictor. Each interaction scores to be freely estimated in children and adolescents
effect was tested separately. In each separate test, all the separately while restraining the other ‘‘common’’ predictors to
predictors including the one whose interaction was being be equal across the two groups. The resulting separate
tested were entered into the regression equation first. The regression coefficients of social self-concept and test scores
interaction term was then entered. An F-test of regression from the two samples, respectively, are also included in
increment was used to determine whether a common or Table 4. As shown in Table 4, social self-concept, which was
different regression slope associated with the concerning almost zero in the child sample, was a robust predictor of life
predictor fit the two age groups, while statistically controlling satisfaction in adolescents. Also as hypothesized, test scores
other predictors. As hypothesised, only the interactions were predictive of life satisfaction in children but not in
involving social self-concept, F ¼ 5.04, p 5 .01, and test adolescents.
scores, F ¼ 5.96, p 5 .01, were significant. Other predictors
did not show significant age-related interaction effects. Discussion
Path analysis was then conducted to test two regression
models. The first was an age-invariant model that excluded the Life satisfaction has traditionally been of research interest in
two age-differential predictors, social self-concept and test adult development. According to J.-P. Leung and Zhang

Downloaded from jbd.sagepub.com at MOUNT ALLISON UNIV on June 21, 2015


INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2003, 27 (2), 182–189 187

Table 4
Age-invariant and age-specific predictors of life satisfaction

Regression coefficient

Nonstandardised Standardised Standard error T-value

Age-invariant predictors
Parental warmth 0.75 0.14 0.35 2.14*
Emotional detachment 1.06 0.30 0.23 4.67**
General self-concept 1.45 0.36 0.33 4.38**
Sports self-concept 0.21 0.06 0.27 0.78
Academic self-concept 0.11 0.03 0.31 0.35
Academic self-concepta 0.78 0.20 0.28 2.73**
Age-specific predictors
Social self-concept
Adolescents 1.06 0.26 0.46 2.29*
Children 0.02 0.00 0.57 0.04
Test scores
Adolescents 1.64 0.12 1.54 1.06
Children 5.23 0.34 1.26 4.14**
a
Academic self-concept and other predictors, except general self-concept, were entered into the
regression equation.
*p 5 .05; **p 5 .01

(2000), only recently have researchers started to examine this mentioned earlier, the apparent lack of public interest in
construct in adolescence and cross-culturally. Few studies have athletic activities as well as the lack of concerted efforts or a
investigated young children’s life satisfaction and, to our system to integrate sport competition into social entertainment
knowledge, the present study represents one of the first to in Hong Kong. Second, academic self-concept was found to
focus on age invariance and differences in explaining life predict life satisfaction equally strongly in adolescents and
satisfaction from a set of school- and family-related self- children. This finding implicitly suggests the influence of a
representations. Our findings that some self-representations cultural tradition that values learning and academic activities as
were age invariant and others were age specific lead to a a major dimension of Chinese socialization. More proximally,
broader abstraction that a person’s subjective affect depends the finding reflects ongoing competitions among students for
on the degree to which domain-specific activities are devel- limited tertiary education placements in Hong Kong. Also true
opmentally salient to the person’s age group. Self-representa- of other countries in the region, resource limitation, together
tions concerning age-relevant activities contribute more to a with a tradition that emphasises social stratification through
person’s life satisfaction whereas those related to activities that schooling, creates a social context where a child’s happiness
cease to be important with development contribute less. What hinges on his/her perceived ability to do well in schools.
is age salient and what is less relevant to a great degree is Finally, test scores were more predictive of children’s than
determined by developmental characteristics that are regarded adolescents’ life satisfaction. This finding can be seen as
as conspecific or universal. For example, the finding that social contradicting a common-sense belief that, as children grow
self-concept was a stronger correlate of life satisfaction for older, schooling and career issues become more relevant.
adolescents than for children in Hong Kong implicitly under- However, in the special social situations represented by Hong
pins a strong universal developmental influence. The hypotha- Kong and, to varying degrees, by other Asian countries as well,
lamus-regulated growth spurt in early adolescence gives a child academic performance can be seen as more relevant when a
never-before-experienced physical and mental capacity and child is younger than when older. The education system in
desire, for example, to socialise outside the family. Responding Hong Kong enforces a strict and elaborate testing sequence. If
to this developmental outcome, the child’s self-system ties self- children do poorly in primary schools, they may not (and
appraisal of social abilities more closely to subjective affect. almost certainly will not) attain test results good enough to be
Hong Kong adolescents in the present study indeed reacted to placed into a good secondary school, which, as mentioned
this developmental change similarly to their Western counter- earlier, is streamed into bands by test scores. Once placed in a
parts (e.g., Richards et al., 1998). In this case, potential good secondary school, however, the odds for getting into
cultural or contextual deviations are deemed secondary to the tertiary education are greatly improved. On the other hand,
valence of a universal developmental trend. children placed into lower-ranked secondary schools have little
However, in other aspects of development, social context chance for a university education. Thus (very unfortunately),
may exert strong influence. Several of the present findings primary school test scores determine not only a child’s
implicitly suggest societal influences on the self-system. First, secondary school placement but whether or not, in the child’s
contrary to the West as well as to other Asian regions (e.g., mind, she will have a ‘‘good’’ and ‘‘happy’’ life.
Dong et al., 1996), sports-related self-concept was not The findings on the association between life satisfaction and
correlated with life satisfaction for either adolescents or family relations render culturally and developmentally consis-
children. In this case, culture or social context provides a tent interpretations. As expected, Steinberg and Silverberg’s
plausible explanation because the finding coincides with, as emotional A/D scale was negatively associated with life

Downloaded from jbd.sagepub.com at MOUNT ALLISON UNIV on June 21, 2015


188 CHANG ET AL. / LIFE SATISFACTION IN YOUNG CHINESE

satisfaction, whereas parental warmth was positively predictive streaming to better understand the contextual effect on
of life satisfaction. These results are in line with other studies of adolescent self-concept.
Hong Kong adolescents (J.-P. Leung & Leung, 1992; Man, This study has several methodological limitations that can
1991), showing the same connection between parent–child be improved in future research to draw more confident
relations and life satisfaction. Similar findings have also been inferences. The age-split sample size was moderate and thus
reported in the West (e.g., Dew & Huebner, 1994). Similarly, potentially attenuates the statistical power. The sampling
the results on the normative comparisons between children and method has also failed to represent the full ability spectrum
adolescents confirm developmental theories derived primarily of the Hong Kong schools. Including school stratification in
from Western research. Children had higher life satisfaction as future research will enable more detailed analyses of the self-
well as higher self-concept than adolescents in most of the system within this particular cultural context. Some of the
domains. This finding is developmentally reassuring in that cross-cultural interpretations have been made by implicit
adolescence is known to be an emotionally stressful period comparisons with the existing literature. Future studies that
(Steinberg & Silverberg, 1986), whereas childhood is blessed include samples from more than one culture can make direct
with positive self-perceptions (Eccles, Midgley, & Adler, comparisons in drawing cross-cultural inferences.
1984). The findings are also consistent with previous studies Manuscript received August 2001
in the West that show declines in self-concepts during the Revised manuscript received August 2002
transition from primary school to middle school (Marsh, 1989;
Wigfield et al., 1991). In Marsh’s study for example, physical,
academic, and appearance related self-concept of Australian References
adolescents all reached the lowest level around Grade 8, which
is equivalent to the age of the present sample. Also, consistent Aiken, L.S., & West, S.G. (1991). Multiple regression: Testing and interpreting
with the present finding on life satisfaction, Western research interactions. Thousand Oaks, CA: Sage Publications.
suggests that adolescents were more critical and self-conscious Arnett, J. J. (1999). Adolescent storm and stress, reconsidered. American
Psychologist, 54, 317–326.
and thus less happy than preadolescents (Simmons, Rosen- Berndt, T.J., Cheung, P.C., Lau, S., Hau, K.T., & Lew, W. J. F. (1993).
berg, & Rosenberg, 1973). A previous study of Hong Kong Perceptions of parenting Mainland China, Taiwan, and Hong Kong: Sex
children in different grade levels (Man, 1991) has also noted a differences and societal differences. Developmental Psychology, 29, 156–164.
Beyers, W., & Goossens, L. (1999). Emotional autonomy, psychosocial
decline in life satisfaction with age. Finally, as hypothesised, adjustment, and parenting: Interactions, moderating and mediating effects.
children scored lower than adolescents on the A/D scale. This Journal of Adolescence, 22, 753–769.
finding supports our interpretation of this scale as one of Block, J., & Robbins, R.W. (1993). A longitudinal study of consistency and
emotional detachment from parents that increases with age. change in self-esteem from early adolescence to early adulthood. Child
Development, 64, 909–923.
Initiated in this study, this scale proves useful for develop- Cournoyer, D.E. (2000). Universalist research: Examples drawn from the
mental studies of pre-adolescence. methods and findings of parental acceptance/rejection theory. In A. L.
Among the normative comparisons between the two age Comunian, & U. P. Gielen (Eds.), International perspectives on human
development (pp. 213–232). Lengerich, Germany: Pabst Science Publishers.
groups, general self-concept and social self-concept did not Dew, T., & Huebner, E.S. (1994). Adolescents’ perceived quality of life: An
show the expected age difference. There are two possible exploratory investigation. Journal of School Psychology, 32, 185–199.
explanations, which again reflect both universal developmental Dong, Q., Weisfeld, G., Boardway, R. H., & Shen, J. (1996). Correlates of social
status among Chinese adolescents. Journal of Cross Cultural Psychology, 27,
trends and contextual influences. First, there are many
476–493.
individual differences in responding to developmental transi- Eccles, J.S., & Midgley, C. (1989). Stage/environment fit: Developmentally
tions or in experiencing transitional stresses (Block & Robbins, appropriate classrooms for early adolescents. In R. Ames & A. Ames (Eds.),
1993). The duration as well as onset of a transitional effect also Research on motivation in education, Vol. 3 (pp. 139–181). San Diego, CA:
Academic Press.
varies among individuals, who were found to show larger than Eccles, J.S., Midgley, C., & Adler, T. (1984). Grade-related changes in the
normal fluctuations in their self-representations during this school environment: Effects on achievement motivation. In J. G. Nicholls
period (Hoge, Smit, & Hanson, 1990). Other research also (Ed.), The development of achievement motivation (pp. 283–331). Greenwich,
CT: JAI Press.
suggests that adolescents rebuild their depressed self-concept
Eccles, J. S., Wigfield, A., Flanagan, C. A., Miller, C., Reuman, D. A., & Yee, D.
and subjective affect at different speeds (Harter, 1998). Thus, (1989). Self-concepts, domain, values, and self-esteem: Relations and changes
it is difficult to time the peaks and troughs of the self- at early adolescence. Journal of Personality, 57, 283–310.
representation system of an age group uniformly. The other Eccles, J.S., Wigfield, A., Midgley, C., Reuman, D., MacIver, D., & Feldlaufer,
H. (1993). Negative effects of traditional middle schools on students’
explanation is sample specific. Allocation to secondary schools motivation. The Elementary School Journal, 93, 553–574.
in Hong Kong is streamed by abilities. The present sample of Erikson, E.H. (1968). Identity: Youth and crisis. New York: Norton.
adolescents was from Band 2 middle schools. This stratum of Fuhrman, T., & Holmbeck, G.N. (1995). A contextual-moderator analysis of
emotional autonomy and adjustment in adolescence. Child Development, 66,
the Hong Kong adolescent population may well enjoy higher 793–811.
than average general self-concept because they are highly Halpern, D.F. (1992). Sex differences in cognitive abilities. Hillsdale, NJ: Lawrence
regarded in this society. However, their academic self-concept Erlbaum Associates Inc.
is not necessarily enhanced and, according to Marsh and Harter, S. (1982). The Perceived Competence Scale for children. Child
Development, 53, 87–97.
Parker (1984), could be more depressed when derived from Harter, S. (1986). Processes underlying the construction, maintenance, and
within-school comparisons that do not benefit from the big- enhancement of the self-concept in children. In J. Suls & A.G. Greenwald
fish-little-pond effect. That is, in this system, a high ability (Eds.), Psychological perspectives on the self, Vol. 3 (pp. 137–181). Hillsdale, NJ:
Lawrence Erlbaum Associates Inc.
child no longer finds himself/herself a ‘‘big fish’’ in a pond with Harter, S. (1990). Causes, correlates, and the functional role of global self-
fish of similar size. However, academic streaming may serve to worth: A life-span perspective. In R.J. Sternberg & J. Kolligan, Jr (Eds.),
group children of similar interests and thus facilitate social Competence considered (pp. 67–97). New Haven, CT: Yale University Press.
interaction. It may be that Hong Kong adolescents have a less Harter, S. (1998). The development of self-representations. In N. Eisenberg
(Ed.) & W. Damon (Series Ed.), Handbook of child psychology: Vol. 3. Social,
depressed social self-concept. Future studies could compare emotional and personality development (5th ed., pp. 102–132). New York:
Hong Kong with other Asian regions that do not adopt ability Wiley.

Downloaded from jbd.sagepub.com at MOUNT ALLISON UNIV on June 21, 2015


INTERNATIONAL JOURNAL OF BEHAVIORAL DEVELOPMENT, 2003, 27 (2), 182–189 189
Heaven, P.C. (1989). Extraversion, neuroticism and satisfaction with life among McBride-Chang, C., & Chang, L. (1998). Parenting styles, emotional
adolescents. Personality and Individual Differences, 10, 489–492. autonomy, and school achievement: Relations to Hong Kong adolescents.
Ho, D.Y.F. (1986). Chinese patterns of socialization: A critical review. In M.H. Journal of Genetic Psychology, 159, 421–436.
Bond (Ed.), The psychology of the Chinese people. Hong Kong: Oxford Morganti, J.B., Nehrke, M.F., Hulicka, I.M., & Cataldo, J.F. (1988). Life-span
University Press. differences in life satisfaction, self-concept, and locus of control. International
Hoge, D.R., Smit, E.K., & Hanson, S.L. (1990). School experiences predicting Journal of Aging and Human Development, 26, 45–56.
changes in self-esteem of sixth- and seventh-grade students. Journal of Petersen, A.C., Compas, B.E., Brooks-Gunn, J., Stemmler, M., Ey, S., &
Educational Psychology, 82, 117–127. Grant, K.E. (1993). Depression in adolescence. American Psychologist, 48,
Hong Kong Equal Opportunities Commission. (1999). Formal investigation 155–168.
report: Secondary school places allocation system. Wanchai, Hong Kong: Richards, M.H., Crowe, P.A., Larson, R., & Swarr, A. (1998). Developmental
Education Department. patterns and gender differences in the experience of peer companionship
Huebner, E.S. (1991). Correlates of life satisfaction in children. School Psychology during adolescence. Child Development, 69, 154–163.
Quarterly, 6, 103–111. Ryan, R.M., & Lynch, J.H. (1989). Emotional autonomy versus detachment:
Huebner, E.S. (1994). Preliminary development and validation of a multi- Revisiting the vicissitudes of adolescence and young adulthood. Child
dimensional life satisfaction scale for children. Psychological Assessment, 8, Development, 60, 340–356.
149–158. Selman, R.L., Levitt, M.Z., & Schultz, L.H. (1997). The friendship framework:
Kagitcibasi, C. (1996). Family and human development across cultures. Hillsdale, Tools for the assessment of psychosocial development. In R.L. Selman, C.L.
NJ: Lawrence Erlbaum Associates Inc. Watts, & L.H. Schultz (Eds.), Fostering friendship: Pair therapy for treatment and
Larson, R., & Richards, M.H. (1991). Daily companionship in late childhood prevention. Modern applications of social work (pp. 31–52). Hawthorne, NY:
and early adolescence: Changing developmental contexts. Child Development, Aldine de Gruyter.
62, 284–300.
Simmons, R., Rosenberg, F., & Rosenberg, M. (1973). Disturbance in the self-
Laursen, B. (1996). Closeness and conflict in adolescent peer relationships:
image at adolescence. American Sociological Review, 38, 553–568.
Interdependence with friends and romantic partners. In W.M. Bukowski, A.F.
Smetana, J. (1995). Parenting styles and conceptions of parental authority
Newcomb, & W.W. Hartup (Eds.), The company they keep: Friendship in
during adolescence. Child Development, 66, 122–128.
childhood and adolescence (pp. 186–210). New York: Cambridge University
Steinberg, L.D., & Silverberg, S.B. (1986). The vicissitudes of autonomy in early
Press.
adolescence. Child Development, 57, 841–851.
Leung, J.-P., & Leung, K. (1992). Life satisfaction, self-concept, and relation-
ship with parents in adolescence. Journal of Youth and Adolescence, 21, 653– Stewart, S.M., Rao, N., Bond, M.H., McBride-Chang, C., Fielding, R.,
665. & Kennard, B.D. (1998). Chinese dimensions of parenting: Broadening
Leung, J.-P., & Zhang, L.W. (2000). Modelling life satisfaction of Chinese western predictors and outcomes. International Journal of Psychology, 33,
adolescents in Hong Kong. International Journal of Behavioral Development, 24, 345–358.
99–104. Stigler, J.W., Smith, S., & Mao, L. (1985). The self-perception of competence by
Leung, Y.W. (2001). The relations of parenting style and academic competence to Chinese children. Child Development, 56, 1259–1270.
early adolescents’ life satisfaction: A longitudinal study in Hong Kong. Unpub- Terry, T., & Huebner, E.S. (1995). The relationship between self-concept and
lished Master’s thesis, The Chinese University of Hong Kong. life satisfaction in children. Social Indicators Research, 35, 39–52.
Man, P. (1991). The influence of peers and parents on youth life satisfaction in Wigfield, A. (1994). Expectancy-value theory of achievement motivation: A
Hong Kong. Social Indicators Research, 24, 347–365. developmental perspective. Educational Psychology Review, 6, 49–78.
Marsh, H.W. (1989). Age and sex effects in multiple dimensions of self-concept: Wigfield, A., Eccles, J.S., MacIver, D., Reuman, D.A., & Midgley, C. (1991).
Preadolescence to early adulthood. Journal of Educational Psychology, 81, 417– Transitions during early adolescence: Changes in children’s domain-specific
430. self-perceptions and general self-esteem across the transition to junior high
Marsh, H.W., & Parker, J.W. (1984). Determinants of student self-concept: Is it school. Developmental Psychology, 27, 552–565.
better to be a relatively large fish in a small pond even if you don’t learn to Yau, J., & Smetana, J. (1996). Adolescent–parent conflict among Chinese
swim as well? Journal of Personality and Social Psychology, 47, 213–231. adolescents in Hong Kong. Child Development, 67, 1262–1275.

Downloaded from jbd.sagepub.com at MOUNT ALLISON UNIV on June 21, 2015

You might also like